Cep Complex Event Processing for Business Process

by : Max J. Pucher



In 2007 announced and shipped a new feature of its Papyrus Process and Content Platform called the UTA User-Trained Agent. This is a no-charge option that adds to its powerful business process and rule functions. What does it offer? The UTA is a machine learning component for CEP. It can be activitated in the scope of a process queue and monitores the tasks in the queue for complex repeating events in real-time. Each task references a large and changing number of elements (defined as objects in the Papyrus WebRepository) making the definition of process rules very time consuming. The UTA however simply monitores its state space for user activity (or other) events and discovers recurring data patterns in the process scope. User performing activity on a process create events that trigger the UTA for analysis. If a pattern is found to be repetitive the UTA will query the user if that is a valid pattern and then recommend this activity as next process step in the future. Actions that do not require user intervention can be set to execute in the background, for example the initiation of a particular process given the reception of a certain data message in relation to a particular data pattern in a business case. Properly authorized supervisors can invalidate training and trigger patterns. Business rules and new elements can be added to the business process at any time without invalidating the training. The UTA will continue to learn from user activity and thus processes are kept up-to-date without the need for process monitoring. Thresholds can be set for SLAs so that tasks that do not trigger events are routed to supervisors. The following supervisor activity is once more transductive training input to the UTA.

ISIS Papyrus is first to deliver a true CEP functionality for business process. It is easy to use and does not require knowledge engineers or huge amounts of computing power. is celebrating 20 years of innovation and the UTA is proof of that!